JP3388556B2 - Method and apparatus for evaluating stationary sound wave - Google Patents

Description

Detailed Description of the Invention

[0001]

[Industrial field of application] Regarding a stationary sound wave evaluation / diagnosis method and apparatus, four parameters (referred to as measurement quantity) of positive peak value, negative peak value, ratio of negative peak value and frequency are used. To do.

[0002]

2. Description of the Related Art The following two methods are generally used as means for evaluating stationary sound waves. (1) Sound level meter (2) FFT analyzer A sound level meter is a simple one that numerically displays the sound volume (strength, power) (unit: "phone") or meter display. F
The FT analyzer measures a time axis waveform of a sound wave and its frequency spectrum while applying a certain band limitation (filtering).

[0003]

Sound waves usually have complicated waveforms, and therefore the evaluation method must be difficult. However, as is clear from the evaluation means described above, the sound level meter is inexpensive, Since the evaluation contents are limited, an expensive FFT analyzer must be used for a slight evaluation of sound waves.

The FFT analyzer is a useful measuring instrument that can analyze the frequency spectrum of sound waves. (A) For stationary sound waves, the spectrum obtained changes when the target section for FFT is different. (B) Since the obtained spectrum is power and does not include phase information, for example, even if the phase of the harmonic of the fundamental wave of the stationary sound wave changes and the timbre changes, it is not reflected in the spectrum. (C) When trying to compare stationary sound waves with spectra, the amount of data is large and the processing becomes complicated. (D) When evaluating and diagnosing the timbre of a stationary sound wave, it is necessary to understand the frequency spectrum data, and to compare and judge by integrating it. Therefore, software processing such as neural network becomes indispensable, and an external computer that runs it is required. Is also needed, and the system becomes large and expensive.

An object of the present invention is to solve these conventional problems and obtain a substantially constant evaluation value if there is a repetition period of several cycles even if the evaluation period of the stationary sound wave is different. Is more economical than the FFT analyzer, that is, -1 The amount of data required for evaluation and the processing operation are large as in the FFT analyzer, and are not complicated. It is an object of the present invention to provide a method and apparatus for evaluating and diagnosing stationary sound waves, which requires less software and hardware than the FFT analyzer.

[0006]

[Means for Solving the Problems]

(1) In the method for evaluating a stationary sound wave according to claim 1, the input stationary sound wave is converted into an electric signal, and the positive peak value V _{+ P} of the signal waveform after cutting the DC component contained in the electric signal and a negative value Of the negative peak value V _-P and the negative peak rate Mr {peak-to-peak value V _PP = V _{+ P} + V _-P relative to the negative peak value V _-P , Mr = V _-P / (V _{+ P} + V _-P )} or the frequency F of the electric signal (these physical quantities are referred to as measured quantities) is measured, and the input stationary sound wave is evaluated by an arbitrary combination of these measured quantities.

(2) In the invention of claim 2, in the stationary sound wave evaluation method according to item (1), an average frequency Fm is used as the frequency F. (3) The stationary sound wave evaluation device according to claim 3 includes a microphone for converting an input stationary sound wave into an electric signal, and a level adjusting means for inputting the microphone output to adjust the level.
A filter for cutting unnecessary frequency components of the output of the level adjusting means, a DC cutting means for cutting a DC component from the output of the filter, a positive peak value V _{+ P} and a negative peak of the output of the DC cutting means. A peak detector for detecting the value V _−P , a means for generating a DC voltage Vf corresponding to the frequency F of the output of the filter or the output of the DC cut means, the output of the peak detector or the Vf
A negative peak rate (the ratio of the negative peak value V _{-P to} the peak-to-peak value V _{+ P} + V _-P ) is calculated from the output of the generating means and the output of the A / D converting means. An arithmetic unit for calculating Mr or the frequency F, the positive peak value V _{+ P} ′, the negative peak value V _−P ′ output from the A / D conversion means, and the negative peak output from the arithmetic unit. Peak rate M
r / frequency F is compared with a reference value to determine whether each is a good / defective, an analysis / determination unit, measurement data of each of the measured quantities, and a good / bad of the analysis / determination unit for the measurement data.
The display / output unit displays the defect determination result or transfers the result to the outside, and the control unit controls the operation of each unit.

(4) According to the invention of claim 4, in the stationary sound wave evaluation device according to item (3), the Vf generating means is configured to output a voltage corresponding to the average frequency Fm of the input signal. It (5) In the invention of claim 5, in the stationary sound wave evaluation device according to (4), the Vf generating means shapes a waveform of an input signal, and a leading edge of an output of the waveform shaper. Alternatively, it is composed of a monostable circuit that outputs a rectangular wave pulse of a predetermined time length at the trailing edge, and an averaging circuit that averages the outputs of the monostable circuit.

(6) In the invention of claim 6, (3) to
In the stationary sound wave evaluation device according to any one of the items (5), the A / D conversion means has a switch circuit, and the positive peak value V _{+ P} and the negative peak value V _-P are input. The DC voltage Vf corresponding to the frequency F or the average frequency Fm is sequentially selected by the switch circuit and A / D converted.

(7) In the invention of claim 7, (3)-
In the stationary sound wave evaluation device according to any one of items (6), the display / output unit displays or transfers each reference value of the measured amount and each allowable deviation together with each measurement data. (8) In the invention of claim 8, in the stationary sound wave evaluation device according to any one of (3) to (7), the reference value or the allowable deviation of any one of the measured quantities or the selection characteristic of the filter is selected. At least one different measurement rank and which measurement rank to apply in each measurement cycle can be set in the analysis / determination unit through the control unit.

(9) According to a ninth aspect of the invention, in the stationary sound wave evaluation apparatus according to the eighth aspect, the display / output unit is
The measurement rank number is assigned to one coordinate axis of the two-axis orthogonal coordinates and the measurement number within each measurement rank is assigned to the other coordinate axis, and whether or not there is a failure in the measurement data is displayed on the corresponding coordinate. There is. (10) In the invention of claim 10, in the stationary sound wave evaluation device according to (8), the display / output unit has a measurement cycle number or elapsed time on one coordinate axis of the biaxial Cartesian coordinate system and an arbitrary number on the other coordinate axis. The measurement amount is taken and whether or not there is a failure in the measurement data is displayed at the corresponding coordinates.

[0012]

【Example】

(1) Stationary sound wave evaluation method In this invention, the four parameters defined by
Of the stationary sound wave using all or part of the
Value, diagnose.   Positive peak value V_{+ P}(> 0) A stationary sound wave is converted into an electrical signal with a microphone and the
Positive of the signal waveform after cutting the DC component contained in the air signal
Maximum amplitude on the side and treated as a positive value.   Negative peak value V_-P(> 0) It is the maximum amplitude on the negative side of the signal waveform described in,
Treat it.   Negative peak rate Mr Peak-to-peak value V of the signal waveform described in_PP= V
_{+ P}+ V_-PNegative peak value V_-PThe ratio of     Mr = V_-P  / V_PP  = V_-P/ (V_{+ P}+ V_-P) (1)   Frequency F or average frequency Fm After converting the stationary sound wave into an electric signal with a microphone,
The frequency F of the reference wave of is measured. Time variation of frequency F
When the frequency is severe, it is desirable to use the average frequency Fm.
Yes. The average frequency Fm is the number of repetitions of the electric signal.
Accordingly, a rectangular pulse with a fixed width τ (its high level is V_H, Low
Level V_LAnd the average voltage Vf is
It is measured and determined by the following equation (2). Invitation of formula (2)
The guide will be described later.

Average frequency = Fm = {1 / (V _H −V _L ) τ} V _f − {1 / (V _H −V _L ) τ} V _L (2) Among these parameters, positive, Negative peak value V _{+ P} , V _-P
Is the magnitude of the sound wave (roundness), and the negative peak rate Mr is a physical quantity associated with the timbre. (2) Stationary sound wave evaluation device FIG. 1 shows the configuration of the invention device 1. Microphone 3
Is used to convert the external stationary sound wave 2 into an electric signal. The converted electric signal is amplified by the amplifier 4 until the sensitivity of the peak detector and the waveform shaper described later is met, and the filter 5
Pass through.

When the stationary sound wave contains a very low frequency component (approximately 1 Hz or less) such as a beat or vibrato, a high pass filter for removing them is used as the filter 5. If you want to remove unwanted high-frequency components, use a low-pass filter. Furthermore, when it is desired to limit to a specific band, a band pass filter is used, or a high pass filter and a low pass filter are used in combination. The control unit 12 controls these filters.

Only the AC component of the output of the filter 5 is taken out through the DC cutting capacitor 6 and input to the peak detector 7. The positive peak detector 7a is, for example, a circuit as shown in FIG. 2A and holds and detects the maximum amplitude during the measurement period.
It is sent to the switch circuit 9a as V _{+ P} (> 0). The negative peak detector 7b is, for example, a circuit as shown in FIG. 2B, and the maximum negative amplitude during the measurement period is held and detected, and V _−P (>
0) is sent to the switch circuit 9a.

The output of the filter 5 is also input to the Vf generating means 8. The Vf generating means 8 is a waveform shaper 8 in this example.
a, a monostable circuit 8b and an averaging circuit 8c.
In the waveform shaper 8a, as shown in FIG.
Is thresholded to perform waveform shaping and output at a specific logic level (for example, TTL level). As shown in FIG. 3C, the monostable circuit 8b outputs a rectangular pulse having a constant width τ for each positive edge of the waveform shaping output. The high level of the rectangular pulse is V _H and the low level is V _L. The averaging circuit 8c is, for example, an RC circuit as shown in FIG. 2C,
The average voltage Vf of FIG. 3C is detected and output.

When the signal frequency is F, there is the following relationship with the signal period T. F = 1 / T (3) On the other hand, since the areas of the V _H level portion and the V _L level portion hatched in FIG. 3C have the same area, (V _H −Vf) τ = (Vf _{-V L) (T-τ)} ··· (4) which is solved for _{T, T = {(V H} -Vf) τ / (Vf-V L)} + τ = (V H -V L) τ / (Vf− _VL ) (5) Substituting the equation (5) into the equation (3), F = (Vf− _VL ) / ( _VH − _VL ) τ ∴F = {1 / (V _H - _VL )?} Vf- _VL / ( _VH - _VL )? (6) is obtained. Here, since V _H , V _L, and τ can be determined in advance, the frequency F can be measured by measuring the average voltage Vf.
Can be asked. It will also be appreciated that if the frequency F has a slight periodic change, the averaging circuit shown in FIG. 2C will provide an averaging effect to measure the average frequency Fm.

The peak values V _{+ P} , V _-P and the average voltage Vf are stabilized after a predetermined time (about 10 times the time constant of the positive / negative peak detector or the averaging circuit) after the input of the stationary sound wave. Therefore, each value is set to A while switching the switch circuit 9a.
It is measured by the / D converter 9b. The control unit 12 switches the switch circuit 9b. The A / D converter 9b measures V _{+ P} , V- _P , and Vf, which are DC voltages. Although the number of bits is 10 to 12 bits, 16 bits or the like may be used when high precision is pursued. These A / D conversion outputs are input to the arithmetic unit 10 and the analysis / judgment unit 11.

The arithmetic unit 10 is composed of, for example, a microprocessor and calculates the negative peak rate Mr and the frequency F of the equation (1) (in this example, the average frequency Fm of the equation (2)).
The analysis / judgment unit 11 is composed of, for example, a microprocessor. The analysis is performed using, for example, the comparison method. Figure 4A
The outline is shown in. Four measured quantities Fm, Mr, V _{+ P} ,
For V- _P , reference values A, B, C and D are first set. The method of setting the reference value is to use the measurement result, use the numerical value input from the operation unit 13, and use the data input terminal 15
There are three ways to use the numerical value transferred from an external device via. Next, the permissible range ± a, ± b, ± c, ± for each measured quantity
Set d. The setting method is for the standard value,
The same method is used.

When the measurement is performed and the measured values A'to D'are obtained, the determination is made for each measured amount. Fm For example, when A−a ≦ A ′ ≦ A + a, it is judged as good (OK), and when other than the above, it is judged as bad (NG).

A parameter (measurement amount) combined by the judgment unit to make a comprehensive judgment based on the analysis result (each judgment result).
What is is designated in advance through the control unit 12. After all, depending on the number of measurands to combine as shown in FIG. 4B,
Any one, two combinations, three combinations, and all four combinations can be classified. For any combination, the total determination is performed on the groups of the number of times of measurement (same as the measurement number) 1, 2, ..., N. If all the measured values are stored in memory, it is possible to perform post-conditional combination conditioning. Judgment is performed in the form of comprehensive judgment, and if there is at least one defect (failure, NG) for each number of measurements, it is defective, and if not, it is pass (good, OK). The analysis / judgment unit 1
1 also controls the display / output unit 14 described below.

The display / output unit 14 displays the analysis / judgment result as an image, notifies by a buzzer or a lamp, and transfers it to an external device. 7A and 7B show numerical display examples. In the figure, REF represents a reference value and MEA represents a measured value. In FIG. 7A, the ranks R1 to R5 are displayed.
The rank represents a measurement unit in which the allowable range of each measurement amount (parameter) and the use of the filter 5 can be independently set. For example, you can use up to R60. In each rank, for example, the number of measurements n
= 1 to 90 data can be stored. When four parameters are displayed on the screen, only five data can be displayed, but the rank directions R1, R2, ..., Rm or the depth directions Ri-1, Ri-2, ... Ri-n of each rank can be freely selected. You can call. In FIG. 7B, the display of only rank R1 is designated. Since the allowable range and the filter condition are common, they are displayed only in one place. Figure 7
In both displays A and B, the comprehensive judgment result of the entire screen display is displayed in the upper left.

FIG. 8A shows that during or after the measurement,
It is a list showing in which measurement NG occurred. Since the coordinate configuration is made up of the rank number Rm and the depth number (measurement number) n, the coordinates of the NG portion from this display can be read and the details can be confirmed on the display of FIGS. 7A and 7B. . FIG. 8B is a display showing changes in four measured quantities with respect to a measurement cycle or elapsed time from the start of measurement. An NG mark and its coordinates are attached to the measured amount that exceeds the allowable range. The maximum number of times of measurement is defined by the number of ranks m and the number of depths n, and the measurement order can be arbitrarily set as described below.

Measurement cycle N: 1 to 5 → 6 to 15 → 16
.. 19 → ..., for example, rank 1 can be performed 5 times → rank 3 10 times → rank 2 4 times → ... The control unit 12 controls general setting and general measurement control.
The operation unit 13 is composed of a keyboard, an encoder and the like, and can manually input setting information and control information. (3) Physical Meaning of Negative Peak Rate Mr Figure 5 shows Electone (YAMAHA PSS-170)
3 is a waveform of the output of the DC cutting capacitor 6 when a sound wave of the trumpet (260 Hz) is input. As shown in the figure, one cycle of the waveform is T (> 0), the positive peak value is V _{+ P} (> 0), and the negative peak value is V _-P (> 0). Also, the area occupied by the positive side waveform at the 0V level is S1, the area between the negative side waveform and 0V is S2, and the remaining area obtained by subtracting S2 from the negative side quadrangle area TV _-P is S3. If the remaining area obtained by subtracting S1 from the square area TV _{+ P on the} positive side is S4, S1 + S4 = TV _{+ P} (7) S2 + S3 = TV _-P (8) Since it is an AC-coupled signal and its average value is 0, S1 = S2 (9) On the other hand, the negative peak rate is Mr = V _-P / (V _{+ P} + V _-P ) (1) Therefore, by multiplying the denominator and the numerator by T, Mr = TV- _P / (TV _{+ P} + TV- _P ) (7) and (8) are substituted: = (S2 + S3) / (S1 + S4 + S2 + S3) (9) substituting equation = (S1 + S3) / ( S1 + S2 + S3 + S4) ··· (10) i.e., the negative peak rate versus the area of the V _PP × T It comes to the ratio of the waveform to the area under that.

This means that if the waveform changes such that the ratio of the lower area changes (of course, the ratio of the lower area does not change, an example of which is shown in FIGS. 6A and 6B), it has a negative peak rate. Indicates that it will be reflected. FIG. 9 shows an example of the case where the scale of the lead organ of the electric crane is changed. The waveform changes, and along with that, the negative peak rate Mr also changes.

On the other hand, a typical change in the waveform without changing the ratio of the lower area is when the waveform is expanded or contracted at a constant rate with respect to 0 V, that is, when the waveform is similar to that of 0 V.
An example thereof is shown in FIG. A has a slight change in the similar waveform due to noise and offset error, and the negative peak rate Mr has increased, but B and C have a high degree of waveform similarity, and the value of the negative peak rate Mr is 48.0, It is in good agreement with 48.7 (%).

From the above, although the waveform cannot be directly estimated from the negative peak rate Mr (for example, the values of Mr and D in FIG. 9A are close to 55.6% and 56.3%, respectively. Is very similar), and it has a certain reference waveform, and it is very convenient to use for detecting a slight change in the waveform. Of course, changes that do not change the proportion of the lower area cannot be detected, but such cases are rare from many experimental results.

In the above description, D is placed on the output side of the filter 5.
C-cut capacitor (generally DC cutting means) 6
Is provided, but the filter 5 is an HPF or BPF.
In this case, since the DC component is cut, the DC cutting capacitor 6 may be omitted. Further, instead of the output of the filter 5, the output of the DC cut capacitor 6 may be input to the Vf generating means 8.

[0029]

As described above, according to the stationary sound wave evaluation method and apparatus of the present invention, even if the stationary sound wave evaluation period is different, a substantially constant evaluation value is obtained as long as there is a repetition period of several cycles. Is obtained. In addition, the negative peak rate Mr
By adopting a new parameter called, it is possible to evaluate the tone color of a stationary sound wave. In the present invention, what is equivalent to learning by the combination of the FFT analyzer and the neural network is provided by a direct method of arbitrarily setting a reference value and an allowable range, and further, there are several level settings (ranks R1 ... Rm and measurement numbers 1 to n can be combined), and an economical evaluation method and device with high cost performance can be obtained. Other effects of the present invention are as follows.

Since the shape of the stationary sound wave (which determines the tone color) is directly expressed by the numerical value of the negative peak rate Mr, more intuitive evaluation can be provided as compared with the method of analyzing the frequency spectrum. Not only the negative peak rate Mr, but also the positive peak value V _{+ P} , the negative peak value V _−P, and the average frequency Fm can be evaluated at the same time, so that evaluation with higher performance can be performed.

The present invention is also suitable for applications such as sound detection of defects or abnormalities in speakers or machine tools in addition to musical instruments.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a circuit diagram showing an example of a main part of FIG.

3 is an operation waveform diagram of a main part of Vf generating means 8 in FIG.

4A is a diagram for explaining a pass / fail judgment method by the comparison method in the analysis / judgment unit 11 in FIG. 1, and FIG. 4B is a combination of various combinations of parameters and measurement times 1 to n in a matrix. FIG. 5 is a diagram showing a state of determining whether or not each element of the matrix is valid.

FIG. 5 is a diagram for explaining the physical meaning of the negative peak rate.
5 is a waveform diagram of the output end of the DC cutting capacitor 6 of FIG.

FIG. 6 is a diagram of two waveforms in which the ratio of the lower area S1 + S3 in FIG. 5 does not change.

7A and 7B are diagrams showing first and second display examples of the display / output unit 14 of FIG. 1, respectively.

8A and 8B are diagrams showing third and fourth display examples of the display / output unit 14 of FIG. 1, respectively.

FIG. 9 is a waveform diagram showing the relationship between the positive and negative peak values V _{+ P} and V _−P and the negative peak rate Mr for each pitch when the electric tone is set to the lead organ mode.

[Fig. 10] In the trumpet mode, the electric tone
60 Hz) is a waveform diagram showing an example of positive and negative peak values V _{+ P} and V _−P and a negative peak rate Mr corresponding to the waveform when the volume knob is changed at a pitch of 60 Hz.

Continuation of front page (56) Reference JP-A-6-117912 (JP, A) JP-A-5-346340 (JP, A) JP-A-60-3527 (JP, A) JP-A-60-262197 (JP , A) JP 5-273266 (JP, A) JP 6-84079 (JP, A) JP 63-85318 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB Name) G01H 3/10 G01H 3/00 G01H 3/06 G01H 17/00

Claims (10)

(57) [Claims] 1. An input stationary sound wave is converted into an electric signal,
Of the signal waveform after cutting the DC component contained in the electrical signal
Positive peak value V_{+ P}, Negative peak value V_-P, Negative peak rate Mr
{Peak toe peak value V_PP= V_{+ P}+ V_-PNegative for
Peak value V _-PIs the ratio of Mr = V_-P/ (V_{+ P}+ V
_-P)}, Or the frequency F of the electrical signal (these physical quantities
V_{+ P}, V_-P, Mr, F are called measured quantities)
The input stationary sound wave is evaluated by any combination of these measured quantities.
A method for evaluating a stationary sound wave characterized by valuing. 2. The stationary sound wave evaluation method according to claim 1, wherein an average frequency Fm is used as the frequency F. 3. A microphone for converting an input stationary sound wave into an electric signal.
Kurohon, Input the microphone output to adjust the level.
Bell adjustment means, The unnecessary frequency component of the output of the level adjusting means is cut.
Filter, A DC capacitor that cuts the DC component from the output of the filter.
Means Positive peak value V of the output of the DC cut means_{+ P}And negative
Peak value V_-PAnd a peak detector that detects The circumference of the output of the filter or the output of the DC cut means
Means for generating a DC voltage Vf corresponding to the wave number F; The output of the peak detector or the output of the Vf generating means is
Means for A / D conversion, From the output of the A / D conversion means, the negative peak rate (peak
Toe peak value V_{+ P}+ V_-PNegative peak value V_-Pof
Ratio) Mr or a calculator for calculating the frequency F, The positive peak value V output from the A / D conversion means
_{+ P}′, Negative peak value V _-P′, Output from the computing unit
The negative peak rate Mr or the frequency F (measurement of these physical quantities
(Referred to as "quantitative amount") is compared with the reference value to determine whether good or bad
An analysis / judgment unit for judgment, Measured data for each of the above-mentioned measured quantities and the measured data
The good / bad judgment result of the analysis / judgment part is displayed.
Is a display / output unit that transfers to the outside, Stationary sound wave including a control unit that controls the operation of each unit
Evaluation device. 4. The stationary sound wave evaluation device according to claim 3, wherein the Vf generating means has an average frequency F of its input signal.
It is characterized by outputting a DC voltage Vf corresponding to m. 5. The stationary sound wave evaluation apparatus according to claim 4, wherein the Vf generating means shapes a waveform of an input signal with a waveform shaper, and a leading edge or a trailing edge of an output of the waveform shaper has a predetermined time length. It is characterized by comprising a monostable circuit for outputting the rectangular wave pulse of, and an averaging circuit for averaging the outputs of the monostable circuit. 6. The stationary sound wave evaluation apparatus according to claim 3, wherein the A / D conversion means has a switch circuit, and the positive peak value V _{+ P} and the negative peak value V The voltage Vf corresponding to the peak value V _−P value and the frequency F or the average frequency Fm is sequentially selected by the switch circuit to be A / D.
It is characterized by converting. 7. The stationary sound wave evaluation apparatus according to claim 3, wherein the display / output unit displays or transfers each reference value of the measured amount and each allowable deviation together with each measurement data. It is characterized by 8. The stationary sound wave evaluation apparatus according to claim 3, wherein at least one of a plurality of measurement ranks having different reference values or permissible deviations of the measurement amount or selection characteristics of the filter is different. And which measurement rank to apply in each measurement cycle can be set in the analysis / determination unit through the control unit. 9. The stationary sound wave evaluation apparatus according to claim 8, wherein the display / output unit takes a measurement rank number on one coordinate axis of the two-axis orthogonal coordinates and a measurement number in each measurement rank on the other coordinate axis. It is characterized in that whether or not there is a failure in the measurement data is displayed at the corresponding coordinates. 10. The stationary sound wave evaluation apparatus according to claim 8, wherein the display / output unit takes a measurement cycle number or elapsed time on one coordinate axis of biaxial Cartesian coordinates and an arbitrary number of the measured amounts on the other coordinate axis. Then, whether or not there is a failure in the measurement data is displayed at the corresponding coordinates.